The rapidly increasing number of antibiotic resistant bacteria is a growing challenge for medicine and health care systems worldwide. Especially, the number of infections with methicillin-resistant Staphylococcus aureus (MRSA) increases dramatically in developed countries. In the Netherlands, hospital acquired infections are managed successfully by screening of patients entering hospital for MRSA carrier and consequent decolonisation of such patients. The established strategy of decolonisation of S. aureus from patients is eradication of S. aureus in nose using an antibiotic (for example, mupirocin) and decolonisation of skin using disinfectants. Problems of this decolonisation regimen are (i) extensive use of antibiotics, (ii) upcoming resistance of S. aureus against antibiotics (for example, against mupirocin), (iii) high costs due to very long treatment time (normally 5-7 days), and (iv) opportunistic infections due to complete eradication of natural skin flora.
Considering these severe disadvantages, the community is seeking for alternative approaches for effective decolonisation within short time and avoiding the use of antibiotics. Lysins, which are bacteriophage-induced lytic or hydrolytic enzymes responsible for bacterial host lysis, offer such advantages in general but they have to be optimized for various aspects. Bacteriophage (phage) lysins have been designated using various names including lysins, phage-lysins, virolysins, and endolysins. Structurally, lysins are commonly found as modular proteins with at least one domain that confers the enzymatic activity for hydrolysing specific bonds in the murein or peptidoglycan layer of the bacterial cell wall (enzymatic active domain—EAD), and one domain that confers binding specificity to a carbohydrate epitope of the cell wall (cell binding domain—CBD). Thus, lysin family members (cell wall or peptidoglycan hydrolases) exhibit a modular design in which a catalytic domain is fused to a specificity or binding domain.
During the bacteriophage (phage) reproduction cycle, after assembly of the new phage particles, lysins (endolysins) are produced to destroy the bacterial cell wall. Endolysins can be divided in into five classes according to their cell wall lytic activities: (1) N-Acetylm uraminidases (lysozymes), (2) Endo-β-N-acetylglucosaminidases, (3) Lytic transglycosidases, (4) N-acetylmuramoyl-L-alanine amidases, and (5) Endopeptidases. While the aforementioned (1) to (3) all cleave within the sugar moiety of the peptidoglycan, (4) cleaves at the amid bond between the sugar backbone and the peptide linker and (5) cleaves within the peptid cross bridge.
Endolysins are described as having a narrow spectrum regarding their target. In case of Gram-positive bacteria, endolysins act on cell walls from inside as well as from outside, thus, making these molecules to antimicrobial drug candidates (Borysowski et al. 2006). While bacteriophage host ranges are largely restrictive, i.e. recognizing only one specific antigen on its bacterial host, phage lysins are less restrictive, recognizing a specific carbohydrate molecule common to the particular species of host bacteria. Although the range of bacteria targeted by lysins is less restrictive than the corresponding bacteriophage, lysins still maintain a degree of specificity, having minimal effects on other bacteria including commensal organisms.
The use of endolysins to kill bacteria was disclosed for the first time by Gasson in 1991 (GB 2 255 561). Further therapeutic and prophylactic applications, including animal model systems, have been described by Nelson et al. 2001. This work describes a topical application of endolysins against group A streptococci and pneumococci in oral and nasopharyngeal treatment. In the field of staphylococcal treatment with bacteriophage derived lysins, Rashel et al. 2007 have shown that endolysin from phage phiMR11 is able to eradicate MRSA in nares of mice and protects mice by intraperitoneal injection from septic death. Further regimes of treatment and pharmaceutical compositions to treat and prevent bacterial infections using phage derived lysins are described in U.S. Pat. No. 5,997,862. However, in all so far published examples using bacteriophage derived endolysins for the treatment of bacterial infections, the amount of protein for an effective treatment is very high. This is due the poor stability of the enzymes and due to inhibition of the activity in application relevant matrices.
In case of lysins against Staphylococcus bacteria, a number of wild-type endolysins have been cloned and characterized. For example, protein 17 from phage P68 is a staphylococcal endolysin, which is reported to exhibits antimicrobial activity against S. aureus isolates including clinical isolates (Takác and Bläsi 2005). Various groups investigated the endolysin of S. aureus bacteriophage phi11 in antimicrobial applications. Navarre et al. 1999 identified two enzymatic activities domains (amidase and endopeptidase) in phi11 lysin and showed that a mutant with deletion of the amidase domain is still active. Mutants of phi11 (and phi12) endolysin have been characterized by different activity assays on S. aureus cell walls, heat inactivated bacteria and on bacterial biofilms (Sass and Bierbaum 2007). All these investigations have in common that they are using artificial experimental conditions for functional characterization of the endolysins. Therefore, no evidences regarding efficacy on living cells under application-relevant conditions can be drawn from these publications.
Another staphylolytic enzyme is derived from bacteriophage phiK. This endolysin, called lysK, has been characterized in more detail by the groups of David M. Donavan and R. Paul Ross (O'Flaherty et al. 2005; WO 2008/001342; Becker et al. 2008; Horgan et al. 2009). They have been able to show, that lysK has a broad bactericidal activity against living Staphylococcus bacteria without discriminating between the different genera. LysK consists of one CBD and two EADs, a cysteine-histidine amino peptidase (CHAP) and an amidase domain. Expressing the individual EAD's, they were able to show that the CHAP domain alone is sufficient for killing but not the amidase domain. A deletion mutant, without amidase domain (lysKΔ221-390), possesses the same killing activity as the wild type protein. Determining MIC values for the truncation/deletion constructs, only MIC values for the wild type LysK and the lysKΔ221-390 were measurable in TSB medium. The CHAP domain alone showed no measurable activity within such a complex matrix. The determined MIC values are considerably high, 78 μg/ml and 63 μg/ml for wild type lysK and lysKΔ221-390, respectively. No chimeric lysin based on lysK domains has been described so far.
All published data using wild type endolysins clearly show that these molecules are quite effective in killing bacteria in buffer solutions. The advantage of these molecules is the very fast onset time (minutes to hours), and the mode of action from outside without involvement of metabolic processes within the cell. As a matter of fact, for endolysins induction/acquisition of resistance has not been described in literature. On the other hand, wild type endolysins tend to be quite unstable at elevated temperatures and functionality is reduced in complex compositions like culture media or biological fluids. All published MIC values (minimal inhibitory concentration) or MBC values (minimal bactericidal concentration) are in the range>50 μg/ml. It can be speculated that in many cases MIC values are not reported for experimental reasons.
Enzymes with cell wall degrading properties similar to bacteriophage lysins (endolysins) can also be found in bacteria. Autolysins are bacteriolytic enzymes that digest the cell-wall peptidoglycan of the bacteria that produce them. Autolysins are involved in cell wall reconstruction during bacterial cell division. Although potentially lethal, autolysins appear to be universal among bacteria that possess peptidoglycan. “Autolysin” is the term used for lysins, which are produced by bacteria and involved in cell division, while the term “lysin” or “endolysin” refers to lytic enzymes, which are involved in phage release, as described herein above. Bacteriocins are molecules also produced and secreted by microorganisms. They are antibacterial substances of a proteinaceous nature that are produced by different bacterial species. A subclass of bacteriocins consists of enzymes (proteinaceous toxins) which are produced by bacteria to inhibit the growth of similar or closely related concurrence bacterial strain(s) in their habitat. Many bacteria produce antimicrobial bacteriocin peptides. They also contain CBDs and EADs. Bacteriocins target prokaryotes but not eukaryotes, making them safe for human consumption.
The bacteriocin lysostaphin is naturally produced by Staphylococcus simulans to combat Staphylococcus aureus. It is highly effective in vitro and capable of killing bacteria in complex media (Kumar J. 2008). Lysostaphin consists of one CBD and one glycyl-glycine endopeptidase domain, which cleaves the characteristic penta-glycine cross bridge in S. aureus cell walls. This molecule has been tested in various animal models and exhibit good efficacy even in complex matrices (Kokai-Kun et al. 2007; Kusuma et al. 2007). The reported MIC values of lysostaphin are more than 1000-fold lower compared to lysK (<0.02 μg/ml). The major disadvantage of lysostaphin is the occurrence of resistance in S. aureus. Two different genetic escape mechanisms have been described so far: First, incorporation of serine into the penta-glycine bridge (DeHart et al. 1995). Secondly, shortening of the glycine bridge; gly3 or gly2 (Ehlert et al. 1997; Strandén et al. 1997). It can be assumed that such monogenic resistance marker will rapidly be selected under selection pressure.
Enzymatic active domains (EADs) can further be found in structural bacteriophage proteins. They are part of the early infection machinery of the bacteriophage, locally hydrolyzing the cell wall prior to DNA injection.
In order to deal with the fact of resistance development, groups started to investigate the combination of different lysins. For example, synergistic effects between lysK and lysostaphin (Becker et al. 2008) have been described, resulting in reduced effective concentrations for killing S. aureus. The drawback of this concept is, that in case of occurrence of resistance against one component (for example, lysostaphin), the concentration of the second component will not be effective anymore. Furthermore, a composition with two active components is difficult to develop and expensive in production.
It is known that a combination of domains (CBD's and EAD's) from different source organisms is possible. However, the purpose of such domain exchange experiments was always to alter or broaden the host specificity of the lysins (Diaz et al. 1990; Croux et al. 1993; Donovan et al. 2006). So far, no systematic domain exchange experiments have been performed with endolysin-derived EAD's to obtain lytic molecules with improved properties with respect to efficacy, resistance potential and stability.
There is one example in the literature to construct a highly stable chimeric lysin based on a lysostaphin CBD fused to a tail associated murein-degrading enzyme (TAME) domain (WO 2007/130655). This domain can be considered as stable as it is a part of a bacteriophage structural protein. The disadvantage of such constructs is the very low specific activity compared to endolysins. Therefore, more protein is required to reach effective concentrations. Furthermore, inhibition of the molecules in complex matrices cannot be excluded because no characterization in this regard has been provided.
There is an ongoing need for therapies and agents effective in the control of bacterial contamination, colonization and infection. A major problem in medicine has been the development of drug resistant bacteria as more antibiotics are used for a wide variety of illnesses and other conditions. The over utilization of antibiotics has increased the number of bacteria showing resistance. Furthermore, broadly reactive antibiotics can affect normal flora and can cause antibiotic resistance in these organisms because of the frequency of drug use. The number of people becoming hyper allogenic to antibiotics appears to be increasing because of antibiotic overutilization. Accordingly, there is a commercial need for new antibiotics (or bacterial killing substances), especially those that operate in new modalities or provide new means to kill pathogenic bacteria.